Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2004-10-14

00:16:05 Quit gromit`` (Read error: 110 (Connection timed out))
00:16:08 Join gromit``` [0] (
00:17:29 Join iRiverMan [0] (
00:21:20_LucretiaAnyone know anything about CalmRisc? it's the CPU inside the MT-500, by the looks of it
00:30:13 Quit zonk ("sudden death")
00:40:37 Quit mecraw__ ("Trillian (")
00:53:29 Quit iRiverMan ("CGI:IRC (Ping timeout)")
00:56:31***Saving seen data "./dancer.seen"
01:02:45 Join Sebulba02 [0] (
01:20:43 Quit AciD (Remote closed the connection)
02:28:04 Part amiconn
02:56:35***Saving seen data "./dancer.seen"
03:10:54 Quit midk (Remote closed the connection)
03:18:00 Join midk [0] (
03:20:25 Join midk_ [0] (
03:20:25 Quit midk (Read error: 104 (Connection reset by peer))
03:20:55 Nick midk_ is now known as midk (
03:47:47 Join knoppix [0] (
04:56:37***Saving seen data "./dancer.seen"
05:50:18 Join MoGle3 [0] (
05:51:31 Quit scott666_ ("i'll be back...eventually...")
06:10:19 Quit gromit``` (Read error: 110 (Connection timed out))
06:11:05 Quit MoGl53 (Read error: 110 (Connection timed out))
06:45:16 Join LinusN [0] (
06:56:39***Saving seen data "./dancer.seen"
07:19:41 Join gromit` [0] (
07:29:57Mode"#rockbox +o LinusN " by ChanServ (ChanServ@services.)
07:32:18Topic"Rockbox - rocks your box" by LinusN (
07:43:37 Join webguest84 [0] (
07:44:56 Quit webguest84 (Client Quit)
08:11:59 Join einhirn [0] (
08:14:26 Join Zagor [242] (
08:16:59 Join amiconn [0] (
08:17:31amiconnGood morning
08:18:41amiconnLinusN: The fake poweroff works nice on Ondio :)
08:19:34LinusNi and zagor discussed it yesterday, and we have pretty much agreed on removing the "shutting down" splash alltogether
08:20:16amiconnHmm. Imho this adds a bit of responsiveness, so I wouldn't remove it
08:20:34LinusNand just fake a poweroff and do the housekeeping when it's "off"
08:22:17amiconnAlthough that should give the same responsiveness, it may irritate the user if he tries to switch back on too early, or simply by the fact that the hd based models may still show some activity after the display is apparently showing "off"
08:23:20amiconnI got the "switch back on too early" problem myself on the player, even with the current rockbox behaviour. I had to learn that I have to wait several seconds before switching it on again
08:23:47*LinusN feels the "off-but-not-really-off" discussion coming up, similar to the dreaded charging screen
08:24:43Zagori say we try it. if it doesn't fell good, we revert.
08:25:24amiconnA propos - the archos recorder fw does exit to the charging screen after the idle timeout ;)
08:25:38LinusNit does?
08:26:20LinusNi guess they have their reasons
08:26:42LinusNthey don't shut of the hard drive power
08:26:49LinusNonly in the charging screen
08:27:08LinusNso they probably charge slightly faster in the charging screen
08:27:18amiconnAbout the shutting down splash - it is displayed for ~1 sec even on Ondio
08:27:31LinusNwhich reminds me that we might want to do the same for the player
08:27:42LinusNamiconn: yes, sleep(HZ)
08:29:17LinusNperhaps the splash is unneccessary on the ondio
08:30:12LinusNi personally think that i'd like to know why the disk spins up when i turn off the fm recorder, so the splash is not entirely useless
08:30:48LinusNfaking a poweroff and spinning up the disk afterwards might be perceived as a bug
08:30:55amiconn(Ondio) I agree. If the ~1 s is caused by the sleep(HZ) only, it makes no sense.
08:31:08LinusNat least the user has no idea what is going on
08:31:27LinusNit could even say "saving settings"
08:31:36LinusNthen there is no confusion
08:31:41amiconn(Recorder) yes, thats what I was trying to say with ... may irritate the user ... by the fact that the hd based models may still show some activity after the display is apparently showing "off"
08:32:35LinusNon second thought, i think the splash should stay, saying "saving settings"
08:33:43 Join AciD [0] (
08:33:46amiconnRemember that rec v1 and player don't switch off immediately with the poweroff() function, because it is done by holding the off/stop key "pressed"
08:35:23LinusNv1 and player can only be cleanly shut off with the OFF "double click"
08:36:13LinusNand the poweroff() function shuts them off instantly
08:37:39LinusNelse the idle poweroff wouldn't work
08:39:26amiconnYes I know. I mean, the splash should be displayed on v1/player even if there is no spinup necessary on shutdown, because the display won't go away immediately.
08:40:00 Quit AciD (Read error: 104 (Connection reset by peer))
08:40:02LinusNi don't understand what you mean
08:40:38 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
08:41:03LinusNif nothing needs to be saved, the shut down is instant
08:41:53LinusNah, now i get it
08:42:07LinusNsorry, me silly
08:42:33amiconnV1 and player don't switch off immediately with the poweroff() function, because there is no real soft poweroff by fliping a port pin, but they do the off by faking a continous press of the off button, and then the hw poweroff kicks in, after ~1 sec
08:42:51LinusNyou are correct
08:44:19amiconnOf course it's power_off(), and I should know, as I fiddled with power.c yesterday ;)
08:56:43***Saving seen data "./dancer.seen"
08:58:08 Join AciD [0] (
09:09:04 Join einhirn [0] (
09:29:47*Bagder noticed
09:32:26 Join Alcoolya [0] (
09:32:49 Quit Alcoolya (Remote closed the connection)
09:33:06knoppixHas anyone reported anything strange about "insert into playlist" / "queue"?
09:33:22 Nick knoppix is now known as kurzhaarrocker (
09:33:43LinusNnot really, why?
09:34:12kurzhaarrockerWhen I try to insert / queue the playlist changes entirely and the newly inserted file is played instantly
09:34:51kurzhaarrockerAfter the inserted file has been played it doesn't return to the original playlist but continues to play the directory where the inserted file is seated.
09:35:11Zagorsounds like it does a simple PLAY instead of insert
09:36:31LinusNis the playlist playing when you do this?
09:37:04LinusNis it a regular playlist or a dir play?
09:37:14Zagorappears to be a button bug
09:37:17Zagori get it too
09:37:30kurzhaarrockerIn my case it was a regular playlist that was resumed from a bookmark.
09:38:05LinusNprobably a stray release event
09:40:09Zagorfound it, fixing
09:40:41 Quit _Lucretia (Read error: 110 (Connection timed out))
09:40:53*kurzhaarrocker attaches a "Incredibly Quick Bug Squisher" award to Zagor
09:42:15*LinusN attaches a "most bugs introduced in one CVS commit" award to Zagor
09:42:22 Join ashridah [0] (ashridah@
09:42:37Zagoryour 1254 warnings is still a record that's hard to beat
09:42:57LinusNthat was one bug that gave many warnings, that doesn't count :-)
09:43:08*LinusN hides
09:43:10ashridahfrom one file?
09:43:24ashridahor was it a file that got included in lots of others...
09:43:46 Join _Lucretia [0] (
09:44:09kurzhaarrockerLinusN you probably are just not as good as Zagor at _hiding_ bugs
09:44:29LinusNhe's sneaky
09:45:02Bagderlet's burn him!
09:45:23Bagderhe turned me inte a newt!
09:45:37kurzhaarrockerbtw: I made my external charger measure the capacity of my batteries once more. They are better than I remembered. Each one has > 1350 mAh.
09:47:31LinusNthen you might have a bad battery connection
09:47:32Zagorand that, my liege, is how we know the Earth to be banana shaped
09:53:16amiconnZagor: Another button bug: If you enter the playlist viewer from the menu, it immediately triggers "play" on the first entry, effectively restarting the playlist (at least on Ondio)
09:53:44Zagorah, haven't fixed playlist viewer. will check.
09:54:19amiconnBtw, I used some of your dynamic button mechanisms for the plugins.
10:00:01amiconnE.g. in the jpeg viewer, zoom in is short press of menu button, while zoom out is long press
10:00:07amiconn..on Ondio
10:02:06kurzhaarrockerBagder: What do I read? Song db support?
10:03:24Zagorkurzhaarrocker: we've started looking at it. it's a lot of work left.
10:04:17amiconnAlthough it may be of use for some users, I'm pretty sure Iwon't use it
10:05:12Zagormy slimp3 has both, and I find I use the database almost exclusively for searching
10:07:42amiconnUsually I know where to find the track I'm looking for. In the rare case where I don't know, the search plugin is sufficient (for me)
10:08:40kurzhaarrockerFor me the database becomes interesting as soon as it containst statistical data like how often I listened to that song, how often I skipped it...
10:09:07Zagorwe don't plan to add that, at least not initially. the database will be read-only.
10:10:51kurzhaarrockerOnce it is established it might be a piece of cake to add that functionality. But still there are more important features lacking.
10:12:04LinusNwhy would the statistics have anything to do with the id3 database?
10:13:12LinusNkurzhaarrocker: for the record, what features do you think are lacking?
10:13:35kurzhaarrockerBecause it would be interesting to search eg for the least often played songs in order to clean them out.
10:14:01kurzhaarrockerAn id3 tag editor that could be used while recording would be great.
10:14:02LinusNkurzhaarrocker: yes, but why mix it with the id3 database?
10:17:18amiconnLinusN: Think id3 database -> song database, then it makes sense. Separating it would just add another database...
10:22:05amiconn@devs: It should be perfectly possible to deal with a fairly large on-disk database with the archos' limited CPU power and RAM. On the Amiga, there is a GUI program that handles the Internet Movie database (several 100 MB) in a proprietary on-disk database format, with reasonable RAM consumption, and not too long response times. Unfortunately it is not open source.
10:23:15BagderI have no doubts that we'll manage
10:23:23amiconnThe program does create that proprietary database format from the imdb text files (also several 100 MB) within a couple of hours
10:23:40amiconn..on a real Amiga, that is
10:24:04Bagderdo any such still exist? B-]
10:24:05amiconnA song database will be much smaller
10:24:22amiconnBagder: yes of course
10:24:29BagderI know, just being mean
10:24:44Bagderwe were amiga users once too you know
10:24:52*kurzhaarrocker even has 1 1/2 C64 :)
10:56:45***Saving seen data "./dancer.seen"
11:02:51 Join plok [0] (
11:07:00 Join Headie [0] (
11:12:21 Quit plok ("I'm outta here!")
11:24:50 Quit _Headie (Read error: 110 (Connection timed out))
11:28:16 Quit _Lucretia (Read error: 60 (Operation timed out))
11:40:48 Join [IDC]Dragon [0] (
11:40:51 Join _Lucretia [0] (
11:41:44[IDC]Dragongood morning folks
12:06:11amiconn[IDC]Dragon: Morning ;)
12:16:14 Join MisticJeff [0] (
12:16:30MisticJeffGood Afternoon
12:20:00LinusNMisticJeff: can you confirm that the H1xx series is discontinued?
12:21:34 Quit Headie (Read error: 54 (Connection reset by peer))
12:27:39Zagorthat would be ironic :)
12:28:23MisticJeffLinusN: sorry i'm here
12:28:52MisticJeffI cannot officially confirm it, but my instincts tell me that it probably has ceased production
12:29:39LinusNhow ironic
12:30:21ZagorI have a hard time believing that iriver can seriously expect to replace the h100 series with the h300. it's way more expensive and caters to another group of people. ipod owners don't necessarily want a photo viewer, especially at twice the price
12:30:43Zagor"ipod" being used in a virtual sense, as "harddisk mp3 player"
12:31:11 Quit kurzhaarrocker (Read error: 104 (Connection reset by peer))
12:32:48ashridahthe h340 is the same price the h140 was here when it hit the stores
12:33:30Zagoralso they have no less than three lines of flash players. why would they cut their hd range in half, unless the product doesn't sell?
12:34:05 Join kurzhaarrocker [0] (
12:34:28Zagorashridah: yes, but prices come down on all models. h140 is much cheaper than h340. why only keep the expensive model, driving price-consious customers to other brands?
12:34:57Zagori think there is something here we are not seeing yet
12:35:01MisticJefflike i said guys i have no official word on that I'm just guessing by what I read
12:35:14Zagoryeah, i'm just speculating and guessing
12:35:18MisticJeffI do know that they are working on a mini-hdd type player
12:35:49MisticJeffand they will be introducing a new 1.8" hdd player in 2005
12:36:03Zagorok, then it makes sense
12:36:21MisticJeffZagor: did you get my email?
12:37:43Zagorstill, look at the massive number of different flash players:
12:38:16MisticJeffoh yeah, iriver holds the market on flash type players
12:39:30Zagorlooks like it. but that also makes me wonder why they would be worried about having three or four hd players, when they have 14(!) flash players
12:40:27Zagoroh well, enough speculating. time will tell.
12:41:41Zagorin any case, the coldfire port is useful even if all h1xx player go off the market tomorrow. coldfire is a common platform and the high-level porting preparations to rockbox have been valuable
12:41:47LinusNtoo bad though, that the H1xx won't be for sale when rockbox is released for it :-(
12:41:58Zagorwe'll see
12:44:24Hadakais H3xx an entirely different platform then?
12:44:47Zagorsurprisingly short life span, if true. iriver announced the ihp-100 on april 17, 2003.
12:44:55ZagorHadaka: no, it uses the same coldfire base platform
12:45:39HadakaZagor: good, thanks
12:46:29LinusNit's not unlikely that they base future players on the coldfire as well, since they have invested quite a lot in the software
12:47:42Hadakawell, as for the H140 vs. H340 price here, H140 is 450 euros and H340 is 530 euros - the difference isn't too great
12:50:38Zagorsure, but it's a very competitive market and not everyone wants to pay a premium for a color screen
12:51:54Zagoririver simply doesn't feel like "one size fits all" company when you look at their flash range, so I didn't expect such an attitude in their hd range
12:52:23LinusNif you can call it a "range" :-)
12:55:32ashridahzagor: given the size of their market, one wonders if they can actually afford to maintain the two series.
12:56:08ashridahsince if you're not making enough items in a run, you're going to be spending more per unit.
12:56:18Zagorbut they can afford 14 flash players?
12:56:48***Saving seen data "./dancer.seen"
12:56:53ashridahZagor: didn't MisticJeff just say they're holding the market there?
12:57:12ashridahthey probably push enough units to justify keeping so many lines available
12:57:38ashridahflash players are probably easier to manufacture and assemble too
12:57:59Zagori doubt it's that simple. but we will see.
12:58:22ashridahbut yeah, discontinuing the h1xx range now seems premature
13:05:54 Join midk_ [0] (
13:12:07[IDC]DragonMisticJeff: How do I place a picture into my forum profile?
13:12:36[IDC]Dragonah, found it, never mind.
13:13:56[IDC]Dragonamiconn: we have an Ondio user reporting that the (Samsung) tuner works
13:14:57amiconn[IDC]Dragon: Yup, already read that :)
13:20:52[IDC]Dragonamiconn: I had no time for further (startup) tests
13:21:56 Quit midk (Read error: 110 (Connection timed out))
13:49:28amiconn[IDC]Dragon: Yesterday I experimented with your startup_io code (after fixing some bugs first).
13:50:11 Part MisticJeff
13:50:26[IDC]Dragonaha, and?
13:50:49[IDC]Dragon(funny video bug, btw)
13:51:23amiconnVideo should work correctly now, if you copy a fresh video file to the Ondio
13:51:38[IDC]Dragonso I did, yes
13:51:48amiconnThe results with startup_io are really weird.
13:51:57[IDC]Dragon(but haven't tested yet, it's rechargeing)
13:52:12amiconnHuh? Recharging?
13:52:48[IDC]Dragonbut I was thinking about a mod to charge from USB
13:53:04[IDC]Dragonwell, a diode and a resistor
13:53:38amiconnWhen the complete IO range is restored prior to RoLo, the archos fw works, independent of what IO values I restore (either those set by archos via Back-Boot, or those set by your bootloader)
13:54:20amiconnWhen IO s not restored, the archos fw consistently hangs when you try to play something, then shuts down after 10 seconds
13:54:49amiconnThe contrast problem no longer exists (probably fixed by the bus controller init)
13:55:09[IDC]Dragonso you say it depends on that *something* is restored, not what
13:55:43amiconnI though I nailed the offending settings, but the I got that strange behaviour - but only *sometimes* it back, of course
13:57:55amiconnThe missing inits seem to be PAIOR=0x5FFF and TSTR=0xE0 (all timers disabled), but sometimes even that does not help...
13:58:34[IDC]Dragonyou hex-edited startup_io.bin offline for that, I guess?
13:58:38amiconnYou could try if you get the strange rolo behaviour too - remember, the archos fw does boot fine without the inits, but fails on playing
13:59:50amiconnNo, I limited the restore-range (the loop in rolo_io_restore() ), and commented out the long*/short* restores
14:00:54[IDC]Dragonsome registers can only be used with certain width, you know
14:01:12amiconnYup. The limited ranges I tested had no such registers
14:06:44amiconnThe PAIOR was wild guessing, because archos boot does that
14:07:05[IDC]Dragonyes, looks familiar
14:08:24[IDC]Dragondoes the writing to certain registers possibly have a positive side effect, like clearing interrupts?
14:09:46amiconnRockbox disables all interrupts before RoLo anyway (IPRA..IPRE are set to 0000)
14:11:01[IDC]Dragon(wild guessing) or producing some edge, for resetting MAS
14:12:10amiconnI wonder whether PA2 is connected to something else than the (nonexistant) tuner power control
14:12:49[IDC]DragonI doubt it
14:13:08[IDC]Dragonbecause the Archos s/w uses it to control the backlight
14:14:15amiconnNo, that's PA14 according to the wiki table (?)
14:14:37[IDC]Dragonah, tuner power, sorry
14:15:34[IDC]Dragonfor the FM-less and Philips models, they could
14:15:52[IDC]Dragonhave you found something in the disasm?
14:16:27amiconnNo, the disasm didn't help me with the inits at all. Perhps I should disassemble the ROM image...
14:16:55[IDC]Dragonbtw, I should control that tuner line properly
14:17:11[IDC]Dragonmust have been pure chance that the tuner was on
14:17:34[IDC]Dragonand not gnerally undesired, for the power drain
14:17:57[IDC]Dragongenerally not desired, I mean
14:18:17amiconnDo you know whether the tuner line is connected for samsung? Because the wiki says it is prepared, but not connected (yours, philips)
14:18:36[IDC]DragonI guess they use it for Samsung
14:18:49[IDC]Dragonthe Philips has a deep sleep mode
14:20:48amiconnThe port pin is input by default with your bootloader. Perhaps it would be a good idea to do the same port inits there that the archos loader does?
14:21:03 Quit kurzhaarrocker ("Trillian (")
14:22:26 Join kurzhaarrocker [0] (
14:27:16 Join Headie [0] (
14:31:38 Join _Headie [0] (
14:33:45 Quit Headie (
14:33:45 Quit _Lucretia (
14:33:45 Quit Sebulba02 (
14:33:45 Quit mbr (
14:33:45 Quit Hadaka (
14:33:45 Quit Ka_ (
14:34:44 Join pillo [0] (
14:34:52NJoinHeadie [0] (
14:34:52NJoin_Lucretia [0] (
14:34:52NJoinSebulba02 [0] (
14:34:52NJoinmbr [0] (
14:34:52NJoinHadaka [0] (
14:34:52NJoinKa_ [0] (
14:35:23 Quit Headie (Read error: 104 (Connection reset by peer))
14:47:28[IDC]Dragonamiconn: there is a difference between the "normal", cutomer-downloadable .ajz, and the flash content
14:47:54[IDC]Dragonthe one in the flash is larger, most likely it contains more inits
14:48:46[IDC]Dragonso I'm not too worried about port inits in my bootloader
14:49:56amiconnYES; I suspected that the flash version does more inits... that's why your loader should perhaps do them too... because one might rolo the archos fw via rockbox (or rockbox shoulddo it)
14:50:13amiconnSorry for Caps'ing...
14:51:06[IDC]Dragonmy loader has nothing to do with Rockbox rolo feature
14:51:21[IDC]Dragonand plenty of code rund inbetween
14:51:55[IDC]Dragonso I see it as Rockbox' responsibility to prepare an environment where the .ajz can run
14:54:45amiconnSounds reasonable. However, remember that there is another problem related to inits - rockbox doesn't run stable from ROM on Ondio, at least for me
14:55:07amiconnAnyway, I should have a look into the archos ROM disasm
14:56:52***Saving seen data "./dancer.seen"
14:57:46 Quit mbr (
14:57:46 Quit Ka_ (
14:57:46 Quit Hadaka (
14:57:46 Quit Sebulba02 (
14:57:46 Quit _Lucretia (
14:58:43NJoin_Lucretia [0] (
14:58:43NJoinSebulba02 [0] (
14:58:43NJoinmbr [0] (
14:58:43NJoinHadaka [0] (
14:58:43NJoinKa_ [0] (
15:06:53[IDC]Dragonamiconn: what's unstable from ROM?
15:18:09amiconnI already reported that several times... Playback stops after some time (several minutes) and rockbox crashes. It does not always happen
15:19:01LinusNcould be a DRAM refresh or wait state problem
15:19:28 Nick midk_ is now known as midk (
15:20:50[IDC]Dragonah, that one
15:21:35amiconnLinusN: I already checked those inits...
15:27:40[IDC]DragonLinusN: are bitfields "polically correct"?
15:28:00[IDC]Dragonthe Philips tuner has a 5 byte packet, grr
15:28:05LinusNwell, they are architecture specific
15:28:15Zagorcompiler specific, even
15:28:35[IDC]Dragonit's for out low level code
15:28:44LinusNi stay away from bitfields
15:28:58LinusNyou mean the philips driver?
15:29:18LinusNthat may have to compile for both sh1 and coldfire
15:29:37Zagor[IDC]Dragon: the thing is bitfields are very loosely defined in the C spec. neither the storage order, size or padding is specified. each compiler may do as it wishes. so it's not good for hardware interfaces
15:29:37LinusNso i wouldn't trust bitfields for that
15:29:42[IDC]Dragonboth gdb, both big endian
15:29:53[IDC]Dragongcc, I mean
15:30:44[IDC]Dragonwhat a pity that this 5 byte info doesn't fit into a 32 bit word
15:31:04[IDC]Dragonit uses only 2 bits from the last byte
15:31:04LinusNso it's not a matter of "political correctness", more portability
15:31:16LinusN2 bits...
15:31:42[IDC]DragonI'm tempted to drop some insignificant ones from the middle
15:32:47LinusNis it important to fit it into a 32-bit word?
15:33:00[IDC]Dragonwould be way nicer to handle
15:33:05LinusNi see
15:34:08[IDC]Dragonit may be OK to hardcode the last byte
15:34:16amiconnMMC has 6-byte packets... without the CRC, that we don't use, there are 5 significant bytes too...
15:34:38 Join webguest72 [0] (
15:35:23[IDC]Dragonanother reuse for my bit placement code, perhaps?
15:35:24 Quit webguest72 (Client Quit)
15:36:09amiconnNope. The bytes have clearly defined functions. 1st is the command, 2..5 contain the parameter (or zero)
15:36:24[IDC]Dragonfor my tuner, I mean
15:37:22amiconnAh, yes. The tuner does use byte fractions for different purposes?
15:37:57LinusNwouldn't the packing/unpacking be an unnecessary overhead?
15:38:42[IDC]Dragon(byte fractions) not really, the PLL is 16 bit in 2 bytes
15:39:08LinusNif you invent your own 32-bit word format, it gets harder for other people to work in the driver code, since it will differ from the data sheet
15:39:28LinusNi'm tempted to yell "KISS"
15:39:40[IDC]Dragonok, as it looks, I will use a 5 byte array
15:40:48[IDC]DragonLinusN: related subject: the tuner has tather 3 layers:
15:41:10amiconn[IDC]Dragon: From a quick glance at the data sheet, it looks like no parameter crosses a byte boundary in an unusual way. So it should be possible to handle this with simple bit masking
15:41:22[IDC]Dragon1) physical interface, which differs per hookup
15:41:42[IDC]Dragon2) logical driver, which could be the same per tuner
15:41:56[IDC]Dragon3) application code, which is tuner-independent
15:42:40[IDC]Dragoncurrently, 2 3 are in the app file
15:43:08[IDC]DragonI think it's beneficial to seperate it
15:43:15LinusNi'm all for the layering
15:43:23[IDC]Dragonso you can recycle 2 for the iriver
15:43:31[IDC]Dragonport 1
15:43:42[IDC]Dragonand have 3 completely different, perhaps
15:44:11amiconn(3) may still depend on tuner features
15:44:24LinusNso we won't use the autotuning in the philips chip
15:44:30[IDC]Dragonlet's hope we find a common ground
15:45:10[IDC]Dragonor 2) has to contain a thread for non-autotuning ones
15:46:28[IDC]DragonI need to switch between 2 tuners at runtime :-(
15:46:48LinusNdoes the philips chip provide some feedback during the autotuning?
15:46:51[IDC]Dragonso, I need a function pointer table or such
15:47:56[IDC]Dragonyou're free to read the status anytime, I'd say
15:48:04 Quit _Headie (Read error: 54 (Connection reset by peer))
15:48:09 Join Headie [0] (
15:49:49[IDC]Dragonplenty of info to read: ready flag, band limit reached flag, PLL frequency, IF counter, stereo indication, level
15:51:43LinusNis there an advantage to doing it ourselves?
15:51:57[IDC]Dragonthe searching?
15:52:04LinusNyes, i mean the code is there for the samsung chip
15:52:27[IDC]Dragondunno, maybe it's quicker
15:53:02LinusNi'd go for the manual mode, makes the application code simpler, as the two tuners look more alike
15:53:20LinusNset_frequency(), read_if_counter()
15:54:40[IDC]Dragonyes, I understand
15:56:18[IDC]DragonOK, so I'll split the radio.c code into app code and tuner code
15:56:44LinusNyou have my blessing
15:56:52[IDC]DragonI need a new CONFIG_TUNER setting
15:56:59LinusNgo ahead
15:57:11[IDC]DragonSamsung, Philips, or to be decided at runtime
15:57:16LinusNdo it
15:57:26[IDC]Dragonok, I'll shut up
15:57:42LinusNyou need runtime, as the mask bits decide
15:58:10amiconn[IDC]Dragon: You could use bitmasks for CONFIG_TUNER...
15:58:21[IDC]DragonI was thinking of that
16:00:41LinusNyou need to be able to select the driver in runtime
16:00:57LinusNso i think your idea with function pointers souns good
16:01:08[IDC]Dragonyes, and the drivers need to know when to compile in
16:01:37[IDC]Dragonor, the SOURCES file
16:01:49LinusNsounds better
16:02:36LinusNlayer 2 could have an array of structs containing the function pointers
16:02:53[IDC]Dragonso, we could have bitflags in CONFIG_TUNER, one bit per tuner
16:03:04LinusNthat array will be 1 item long in the fmrec and iriver case
16:03:42[IDC]Dragonor a tiny "wrapper" that calls the actual function
16:04:06LinusNstruct fmradio *fm = init_fmradio(PHILIPS_CHIP);
16:04:13LinusNor something like that
16:04:27LinusNthen fm->set_frequencey() etc
16:05:07LinusNthe wrapper adds the chip check overhead to each call
16:05:29[IDC]Dragonyou still didn't say if bit flags in CONFIG_TUNER sound like a good idea
16:05:45LinusNi don't know which problem it would solve
16:06:15[IDC]Dragonhaving one #define to setup the compile time modules
16:07:09LinusNi think HAVE_SAMSUNG... and HAVE_PHILIPS... combined with the SOURCE file would do the job nicely
16:07:33LinusNi have to go now
16:07:46[IDC]Dragonin a addition to a general HAVE_TUNER?
16:07:55[IDC]Dragonlooks clumsy
16:08:19[IDC]DragonI'll do something nice, trust me ;-)
16:08:22[IDC]Dragonc u
16:08:22LinusNdo whatever you fell is the simplest
16:08:35 Part LinusN
16:14:36Zagorarchos gmini xs200 released. it's really tiny.
16:14:43amiconn[IDC]Dragon: Is there a way to figure out what chip would fit into the "backlight provision" space on Ondio?
16:18:49dwihnoit's tiny
16:20:16[IDC]Dragonamiconn: Sipex SP4403
16:21:43amiconnSo we could actually mod the Ondio to have a backlight... btw: How do you know?
16:22:38[IDC]DragonGoogling for Sipex EL backlight
16:23:37amiconnAh. The Sipex was guessed?
16:24:18[IDC]Dragonsince it's their favourite company for power
16:24:31 Quit _Lucretia (Read error: 110 (Connection timed out))
16:25:21 Join _Lucretia [0] (
16:25:37[IDC]DragonI have to find a suitable EL foil
16:30:00[IDC]Dragonmaybe the mobile phone modding scene has something
16:31:15dwihnoLucretia Noin!
16:33:09amiconn[IDC]Dragon: Is it possible to cut EL foils into suitable pieces?
16:33:21[IDC]Dragonyes, that's quite common
16:33:42amiconnSegor offers a blue EL foil, 3x25 cm for 14,80
16:33:52[IDC]Dragonbut the ones I've seen are way bigger, with bulky electrodes
16:34:33amiconnThey also offer EL "Wires", which are sold per cm...
16:34:49[IDC]Dragonplus, you can't "share" it, have to use the end with the electrodes
16:35:16 Join mecraw__ [0] (~lmarlow@
16:35:54 Quit ashridah ("sleep")
16:47:41amiconn[IDC]Dragon: Conrad has some smaller EL foils, only 0.5 mm thick
16:48:35[IDC]Dragonwhat size? small electrodes?
16:50:51amiconnThe smaller variant is 138x34 mm. Unfortunately direct linking doesn't work for the Conrad site, search for "Leuchtfolie"
16:51:59Zagorgotta go. bye.
16:52:00 Part Zagor
16:53:27[IDC]DragonI hate their site
16:54:19[IDC]Dragonlooks like rather thick wires
16:54:36[IDC]Dragonbut is cheaper
16:56:25amiconnThe wires need some isolation... it's a rather high voltage
16:56:55***Saving seen data "./dancer.seen"
17:00:02*[IDC]Dragon shudders thinking about how 120V 400 Hz AC may interfere to the microphone
17:01:33amiconnThe sipex datasheet talks about 40..64 kHz
17:02:02[IDC]Dragon"sounds" better
17:02:36[IDC]Dragonwe may need a suitable foil for ultrasonic
17:02:46amiconnNext question is where to get that sipex chip
17:03:31[IDC]Dragonfirst I'd like to check the pins, if this really matches
17:05:21 Quit einhirn (Read error: 54 (Connection reset by peer))
17:10:43[IDC]Dragonvideo looks really bad without backlight, or why the sudden interest?
17:15:23amiconnVideo doesn't look too bad without backlight if the ambient light is coming from the right direction. But then there is that empty space on the PCB... and blue (or white) backlight would look cool
17:16:15[IDC]Dragonbe glad you don't own an FM Recorder, that has an empty space for a RDS decoder
17:16:29amiconnPerhaps I could also get that philips tuner chip, together with the associated passive stuff
17:16:53[IDC]DragonI can send you a MAS
17:17:37amiconnThen I would have to get that mic, and the line in socket....
17:17:45[IDC]Dragonbut for less effort, you'd better buy an FM
17:18:01amiconn;) Still without backlight...
17:18:40[IDC]Dragonso let's focus on the backlight, which money can't buy
17:18:50amiconnI wonder if there are backlight-enabled Ondios on the market at all. Perhaps only engineering samples had it...
17:19:11[IDC]Dragonphoto models
17:20:08amiconnI found the sipex datasheet
17:20:47[IDC]Dragongoogle found it for me, yes
17:22:08[IDC]DragonI thought you had it already, you mentioned the frequency
17:26:11[IDC]Dragonamiconn: how about this:
17:32:46 Join methangas [0] (
17:47:19[IDC]Dragonthe world is small, it can be bought in a swedish webshop
17:47:42[IDC]Dragonhow much is 111,25 Kr. ?
17:49:09[IDC]Dragon12,23 EUR, found it
17:50:08[IDC]Dragonamiconn: did Linus already ship the converter? Else he may order the backlight and add that.
17:53:59 Join webguest72 [0] (
17:59:17[IDC]Dragonthe electrode position of that foil matches the Ondio layout *very* well...
18:02:00 Quit kurzhaarrocker (Read error: 104 (Connection reset by peer))
18:03:06amiconn[IDC]Dragon: Linus told me that he already shipped the converter
18:05:18[IDC]Dragondoesn't that foil look very promising?
18:09:06amiconnYes indeed.
18:24:54 Join _aLF [0] (
18:30:00 Join einhirn [0] (
18:30:06[IDC]DragonI'm out of here
18:30:10 Quit [IDC]Dragon ("CGI:IRC")
18:34:27 Part pillo
18:44:06 Quit einhirn (Read error: 54 (Connection reset by peer))
18:45:38 Quit webguest72 ("CGI:IRC (EOF)")
18:56:56***Saving seen data "./dancer.seen"
19:54:43 Join ichwillrockgrl [0] (
20:01:27 Join MoGl53 [0] (
20:04:18 Part ichwillrockgrl
20:09:52 Quit MoGle3 (Read error: 60 (Operation timed out))
20:56:57***Saving seen data "./dancer.seen"
21:15:23 Join uski [0] (
21:54:15 Join scott666_ [0] (
21:57:23amiconnBagder: r u there?
21:59:31 Join Domonoky [0] (
22:18:49 Quit gromit` ("Client exiting")
22:33:57 Quit Domonoky (Read error: 104 (Connection reset by peer))
22:36:41 Quit methangas (" HydraIRC -> <- \o/")
22:42:17 Join webguest55 [0] (
22:42:51webguest55anyone here
22:43:56webguest55could you help me install the software in a jukebox 6000
22:44:23webguest55i never had a problem when i did my 15 but i cant get the 6000
22:44:23uskifirst, connect your jukebox 6000 to your computer
22:44:23uskitell me when it is connected
22:44:23DBUGEnqueued KICK webguest55
22:44:23webguest55i downloaded the software to the 6000
22:44:37webguest55its connected
22:45:17uskiyou need to get the rockbox version for jb6000
22:45:24webguest55got it
22:45:36uskidownload this to your archox root foler
22:45:42uskiand then, unzip it to the root folder
22:45:49webguest55how do i get it to the root folder
22:46:15uskido you have a tool for unzipping ?
22:46:26uskiwhat tool is it ?
22:46:48webguest55not sure hmmm
22:46:57uskiWinZip (tm) maybe ?
22:47:05webguest55im using mozilla
22:47:06uskior perhaps WinRar ? PowerArchiver ? ... ?
22:47:21uskii don't think mozilla can decompress files from a .zip file
22:47:25webguest55that ok? files are files that contains other files
22:47:34webguest55let me open a IE window
22:47:35webguest55ahhh ok
22:47:39uskiso, mozilla allows you to download the .zip file
22:48:01uskibut not to extract the files that are inside the .zip file :)
22:48:01uskisee what i mean ?
22:48:08uskiso, you need some tool to decompress the .zip file
22:48:27uskiwhat is the version of windows that you are using ?
22:48:43webguest55downloading winzip
22:48:49uskiok ;)
22:48:55uskithen you need some other software (winzip is fine)
22:50:20webguest55i need to extract correct?
22:50:48uskiyou need to extract all the files and directories that are inside the .zip file, to the root folder of your archos
22:51:00uskii.e. if the drive letter of your archox is X, extract it to X:\
22:51:12webguest55now how do i do it to the root folder
22:51:27uskithe root folder is X:\ :)
22:51:52uskiif you have a directory name after the \ then it's not the root folder anymore: it's a subfolder of the root folder
22:52:06webguest55ok im lost
22:52:11uskidoh :)
22:52:13webguest55how do i extract to the root folder
22:52:18webguest55the drives name is f
22:52:27uskithen tell winzip to extract to F:\
22:52:29webguest55do i extract all files into F
22:52:38webguest55now what
22:52:58webguest55i should see all the files when i open the f drive correct?
22:53:00uskitell me what are the files you have on the root folder
22:53:11uskithere must be some "archos.rec" file if i remember correctly
22:54:01uskinow, close this window, and disconnect your archos from the computer as you would do after transferring files to it
22:54:07uskithen, you can turn off the archos, and then turn it on
22:54:15webguest55then what
22:54:15uskiit should then boot into Rockbox :)
22:54:21uskithen you're done :D
22:54:29uskitry now, so I can help you if something is wrong
22:54:31webguest55do i need to run that archos.mod file?
22:54:38uskiit is run automatically by the archos
22:54:39webguest55brb k?
22:54:46uskisure, but don't last too long
22:54:50uskii can't stay for a lot of time
22:54:57webguest55tryin now
22:54:59uskibut you have some minutes ;)
22:56:59***Saving seen data "./dancer.seen"
22:57:26webguest55i think its working
22:59:06webguest55does it play winamp files?
22:59:38uskiwhat do you mean ?
22:59:47uskiwhat are you calling "winamp files" ?
22:59:50uskiplaylists ?
23:02:20webguest55i cant get a song to appear
23:02:39webguest55i dragged and dropped a song on it but it wont show up
23:06:06uskiwhat is the format of the song ?
23:06:10uskiit must be .mp3
23:06:14uskiif it is .wma it will not work
23:06:22uskiwinamp is a player
23:06:32uskiclick on the file with the right button
23:06:35uskithen Properties
23:06:47uskisomewhere you should see something like .wma, .mp3, and so on
23:06:54uskihmm ok
23:06:58webguest55no song info shows up
23:07:04uskithen the file must be a .wma or another unsupported format
23:07:13webguest55whats a supported one
23:07:13uskitry with another song
23:07:16webguest55what software
23:07:23webguest55what software
23:07:27uskiwinamp is a music player
23:07:32uskiit can play alot of file formats
23:07:34uskiincluding mp3
23:07:38uskibut also a lot of others
23:07:40webguest55so how do i change the file format
23:07:48uskiyou need to convert the file
23:07:54uskii am not sure of the way to do it
23:07:57webguest55with what software
23:08:02uskii don't know :\
23:08:29webguest55thanks though!
23:08:53uskinope ;)
23:09:01uskimy paypal id is
23:09:05uskiyou now owe me $500
23:25:55MoGl53just so you know - ill be posting that broken iriver to you this weekend
23:26:22 Join haruki [0] (
23:29:08harukihello, how does one do a timed recording with rockbox, is this possible with setting time split as well as setting a sleep timer?
23:30:22amiconnharuki: There is no timed recording feature in rockbox yet
23:31:29 Quit mecraw__ (Read error: 232 (Connection reset by peer))
23:31:42 Join mecraw__ [0] (~lmarlow@
23:32:58harukithanks amiconn! but would a five hour recording be possible by setting time split to five hours, and then setting a sleep timer for five and a half hours?
23:35:59amiconnTime split would start a new file after 5 hours in this case. I don't know whether the sleep timer would switch off the box while in recording mode. I never used the sleep timer, and rarely use recording either. You could try it, with shorter times of course
23:37:00 Quit haruki (Remote closed the connection)
23:37:48 Join haruki [0] (
23:38:23amiconnJust started a test myself, as I'm curious...
23:38:57 Quit webguest55 ("CGI:IRC (EOF)")
23:39:00harukisorry mozilla crashed :/
23:39:18harukidid i miss anything?
23:39:34harukiok cool, thanks :)
23:39:56harukii'm going to test it too
23:39:57amiconnYou'd have to wait ~15 minutes for my test result (shortest possible sleep timer setting)
23:40:41harukii know, i was just realizing the same thing :)
23:41:53harukido you know if there is any interest in a timed record setting? i was searching the rockbox list archives and didn't see anything (so far)
23:43:57amiconnAfaik there is someone who already implemented timed recording locally, but he says it is not ready for release yet.
23:45:37amiconnIt was mentioned once or twice on irc iirc.
23:50:03amiconnAh, finally found the init problem that caused the instability with cold-started rockbox on Ondio :-)
23:53:04amiconnTimed recording kludge works - box has switched off itself
23:54:40amiconnAs the second recorded file is also there, you can do this even without timesplit. It seems that the recording is properly stopped before the sleep timer switches off the box
23:57:28harukiok cool! booting my recorder now, to see my result (hopefully the same as yours!)
23:58:38harukimy ten minute recording is only nine minutes, did you have this too? perhaps i need to run vbrfix
23:58:47 Quit uski (Read error: 113 (No route to host))

Previous day | Next day